This entire article is utterly ridiculous for a so-called "data-driven" company to put anywhere, even their blog. The PC+N data analysis they use is flawed by

This entire article is utterly ridiculous for a so-called "data-driven" company to put anywhere, even their blog. The PC+N data analysis they use is flawed by an analytical issue Uber never bother to mention: the PC+N "no-show" rates include when the customer is told there's no car available to dispatch. [0] Not only do Uber not include this data in their figures - they really can't, as their app doesn't allow a car to be requested if none is available for the current location. Uber link to the Goldman School of Public Policy analysis of the PC+N data only in random parenthetical in the middle of the article, with a disclaimer that Uber use the PC+N analysis for the "irony." That doesn't sound very "data-driven" to me. [0]: http://www.taxi-library.org/goldman-dispatch-2007.pdf http://www.taxi-library.org/goldman-dispatch-2007.pdf